A simulation to rehearse it, a field notebook to do it, and everything you need to run it, for the transect skill in B4.1.4.
B4.1.4 (Range of tolerance of a limiting factor) carries an application-of-skills statement for both SL and HL. In summary:
So the simulation cannot replace the fieldwork; it sits either side of it. Use it to teach the logic before students go out, and to stress-test their thinking about correlation afterwards. The field notebook is where the real, student-collected data lives.
Hook. Show the tolerance-curve figure (sim section 1). Ask: why would a plant be common in one spot and absent a few metres away?
Predict. Students commit to predictions for the three species in section 2. No changing later.
Sample. Pairs lay one transect along the gradient, then run the challenges below. Circulate and ask each pair to justify their interval.
Analyse & reveal. Pool data, read the kite diagram, test each species with Spearman's rank, then reveal the tolerance curves and discuss the campion and moisture traps.
Plan. Students start section 1 of the field notebook for your real site: variable, species, question, hypotheses.
Brief on site. Walk the gradient together. Point out the extremes. Remind students of ethics and safety.
Fieldwork. Groups of 3: one handles the tape and quadrat, one reads the sensor, one enters data on a phone. Rotate roles each transect. Aim for ≥3 transects per group.
Before leaving. Every group exports or copies its CSV. Data lives only on the device until it's exported.
Graph it. Scatter graph of abundance against the abiotic variable, plus an optional kite diagram by distance.
Test it. Calculate rs by hand using the working table, then check it against the notebook. Compare with the critical value.
Conclude & evaluate. State the correlation, describe the graph's shape, name one confounding variable and one improvement to the method.
| Challenge | What students should discover |
|---|---|
| Lay the tape top-to-bottom, parallel to the woodland edge. | Light hardly varies, so there's no gradient to correlate with. The sim flags it. A transect has to run along the gradient. |
| Set the interval to 6 m. | Red campion lives in a narrow band at the edge and can be missed entirely. Interval must suit the scale of change. |
| Test red campion against light. | A hump-shaped distribution produces a weak or misleading rs. Spearman's rank only detects monotonic trends. |
| Switch the x-axis to soil moisture. | Wood sorrel correlates strongly, but in the model it only responds to light. Canopy makes the ground dark and damp at once: a confounding variable. |
| Test any species against pH. | No gradient, no correlation. A useful null result. |
| Add three parallel repeats. | n rises, the critical value falls, and patchiness averages out. Replication improves reliability. |
The site is regenerated with "New site", so each class gets different numbers but the same biology. The species responses are a teaching model grounded in real habitat preferences, not field data.
Look for a place where one abiotic variable changes clearly over 10–30 m and wild species dominate.
| Site | Gradient | Variable & sensor | Species that often respond |
|---|---|---|---|
| Hedgerow or tree line into rough grass | Shade → open | Light intensity · light meter / logger | Ground ivy, plantains, grasses, mosses |
| Footpath edge into grassland | Trampled → undisturbed | Distance, soil compaction, or soil moisture | Plantains and daisies near path; taller grasses away |
| Pond or ditch margin | Wet → dry | Soil moisture · moisture probe | Rushes, sedges, moss cover |
| Wall or tree trunk | Aspect or height | Light or humidity | Lichens, mosses, algae (% cover) |
| Rocky shore | Low → high shore | Height above low water, exposure time | Barnacles, limpets, wracks, periwinkles |
Spearman's rank correlation coefficient (rs) suits transect data because abundance is often skewed and relationships are rarely linear. It ranges from −1 to +1. If |rs| equals or exceeds the critical value for n, the correlation is significant at p = 0.05 (two-tailed) and H₀ is rejected.
| n | 5 | 6 | 7 | 8 | 9 | 10 | 12 | 14 | 16 | 18 | 20 | 25 | 30 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| rs | 1.000 | 0.886 | 0.786 | 0.738 | 0.700 | 0.648 | 0.587 | 0.538 | 0.503 | 0.472 | 0.447 | 0.398 | 0.362 |
Critical values for p = 0.05, two-tailed. Published tables differ slightly in the third decimal place; match the table your students will meet in class. With n = 5, only a perfect correlation is significant: push for 10 or more quadrats.
Abiotic factors co-vary along gradients. A transect shows correlation; a controlled experiment is needed for causation.
A hump-shaped response (an optimum) defeats a monotonic test. The range sampled may also be too narrow.
Along a transect, placement is systematic at regular intervals; that's what lets you track change along the gradient.
A single line can run through an unusual patch. Parallel repeats sample the gradient more representatively.
Borrowed from the IB teaching community: run a corridor transect with objects standing in for organisms. To keep it a true abiotic correlation, set it up along a real gradient (for example, distance from a window, measured with a light meter) and place the "species" (paper clips, sticky notes, erasers) with more of one type near the light and more of another deep in the corridor. Students run the full method with the field notebook, and the statistics work exactly as outdoors. It's a good rehearsal, but it doesn't meet the requirement to collect data from a natural or semi-natural habitat.